Paperonium

← Back to the plain-language case

Abstract says LabMol-149 has EC50 below 500 nM; Table 3 gives 1450 and 509 nM

PLoS Comput Biol (2020) · doi:10.1371/journal.pcbi.1007025 · PMC7048302

Summary

The Abstract reports that two of the five compounds chosen by the paper's deep-learning screen, LabMol-149 and LabMol-152, showed antiplasmodial activity "at low nanomolar concentrations (EC50 <500 nM)". The final paragraph of the Discussion says both compounds "completely satisfy" a criterion of EC50 <500 nM against sensitive and resistant strains. The paper's own Table 3 gives LabMol-149 EC50 values of 1.450 ± 0.524 μM against the sensitive 3D7 strain and 0.509 ± 0.177 μM against the resistant W2 strain. These are 1450 nM and 509 nM, so the compound is above the stated threshold on both strains, by a factor of 2.9 on 3D7. Only LabMol-152 (49 and 78 nM) meets the criterion. The headline count of qualifying compounds is therefore one, not two.

What the paper reports

Abstract:

Among them, two compounds, LabMol-149 and LabMol-152, showed potent antiplasmodial activity at low nanomolar concentrations (EC50 <500 nM) and low cytotoxicity in mammalian cells. Therefore, the computational approach employing deep learning developed here allowed us to discover two new families of potential next generation antimalarial agents, which are in compliance with the guidelines and criteria for antimalarial target candidates.

Author summary:

This allowed us to discover two new potential antimalarial candidates.

Discussion, final paragraph:

Moreover, two of these compounds (LabMol-149 and LabMol-152) completely satisfy the guidelines and criteria for discovery of new antimalarial drugs, i.e., activity at low nanomolar concentrations (EC50 <500nM) against sensitive and multiple resistant strains of Plasmodium spp. and SI greater than 10 folds [12,42].

Table 3, header and LabMol-149 row (columns: compound | EC50 (Pf3D7)(μM) | EC50 (PfW2)(μM) | CC50 (COS-7) (μM) | SIa | SIb):

LabMol-149 | 1.450 ± 0.524 | 0.509 ± 0.177 | >100 | >71.4 | >196

Table 3, the other two active candidate compounds:

LabMol-151 | 1.911 ± 0.292 | 1.616 ± 0.321 | 140.8 ± 11.2 | 73.7 | 87.1

LabMol-152 | 0.049 ± 0.029 | 0.078 ± 0.014 | 16.7 ± 10.3 | 340.8 | 214.1

Table 3 footer:

SIa: Selectivity index calculated by COS7 CC50/3D7 EC50; and SIb: Selectivity index calculated by COS7 CC50/W2 EC50; The data are expressed as mean ± SD of three independent assays.

Two related passages use different wording. Results, "Experimental validation":

The EC50 for each compound (Table 3) indicate that two compounds, 2-(4,6-diphenyl-1,2-dihydro-1,3,5-triazin-2-yl)phenol (LabMol-149), 4-{N-[3-(morpholin-4-yl)-1,4-dioxo-1,4-dihydronaphthalen-2-yl]acetamido}benzoic acid (LabMol-151) and N2-(3-fluorophenyl)-N4-[(oxolan-2-yl)methyl]quinazoline-2,4-diamine (LabMol-152), were potent at inhibiting the parasite growth, showing activities in submicromolar and low nanomolar range against both 3D7 and W2 strains.

Discussion, third paragraph:

Two compounds (LabMol-149 and LabMol-151) showed activity at submicromolar concentrations against asexual blood stages and low cytotoxicity in mammalian cells.

What the paper's own numbers show

Method. I converted every Table 3 EC50 from μM (the unit in the column header) to nM and compared each one with the 500 nM threshold stated in the Abstract and the Discussion. The Discussion applies that threshold to both sensitive and resistant strains, so a compound qualifies only if both its 3D7 and W2 values are below 500 nM. For LabMol-149 I also looked at mean − 1 SD, using the footer's statement that the values are mean ± SD of three assays. I recomputed every selectivity index from the footer's definitions (CC50/EC50), and checked the "submicromolar" (below 1 μM) wording against the same values. Every number below is printed by recompute.py.txt.

Quantity Location Reported Recomputed from the paper's numbers
LabMol-149 EC50, 3D7 Table 3; Abstract and Discussion say <500 nM 1.450 ± 0.524 μM 1450 nM, 2.90 × the 500 nM threshold; mean − 1 SD = 926 nM
LabMol-149 EC50, W2 Table 3; Abstract and Discussion say <500 nM 0.509 ± 0.177 μM 509 nM, 1.018 × the threshold (9 nM above it)
LabMol-152 EC50, 3D7 / W2 Table 3 0.049 / 0.078 μM 49 / 78 nM, both below 500 nM
LabMol-151 EC50, 3D7 / W2 Table 3 1.911 / 1.616 μM 1911 / 1616 nM, both above 500 nM
Tested LabMol compounds with EC50 <500 nM on both strains Abstract; Discussion, final paragraph two (LabMol-149 and LabMol-152) one (LabMol-152)
LabMol-149 SIa Table 3 >71.4 100 / 1.450 = >69.0; 71.4 corresponds to an EC50 of 1.401 μM
LabMol-149 SIb Table 3 >196 100 / 0.509 = >196.5
LabMol-151 SIa / SIb Table 3 73.7 / 87.1 73.7 / 87.1
LabMol-152 SIa / SIb Table 3 340.8 / 214.1 340.8 / 214.1
Chloroquine SIa / SIb Table 3 4,545 / 289 >4545.5 / >289.0
Pyrimethamine SIa / SIb Table 3 2,702 / 5.5 >2702.7 / >5.5
LabMol-149 "submicromolar" Discussion, third paragraph submicromolar 3D7 1450 nM, not below 1 μM; W2 509 nM, below 1 μM
LabMol-151 "submicromolar" Discussion, third paragraph submicromolar 1911 / 1616 nM, below 1 μM on neither strain

The table's arithmetic is internally consistent. Every selectivity index recomputes from the printed EC50 and CC50 values, except LabMol-149's SIa, which is slightly high (71.4 against 69.0). The problem is the text. It counts LabMol-149 among the compounds below 500 nM, and the table does not support that for either strain. The Results passage also says "two compounds" and then names three.

Figure

The Abstract and Discussion claims set against the Table 3 values for LabMol-149 and LabMol-152

Figure 1. On the left (orange), the two passages that state the EC50 <500 nM criterion and name LabMol-149 and LabMol-152 as meeting it, quoted from the Abstract and the final paragraph of the Discussion. On the right (blue), the Table 3 values for the same two compounds, converted from μM to nM. LabMol-149 is 1450 nM against 3D7 (2.90 times the threshold) and 509 nM against W2 (1.02 times). LabMol-152 is 49 and 78 nM. Only LabMol-152 meets the criterion as the text states it.

EC50 mean ± SD for the three active candidates on a log axis, with the thresholds the text claims

Figure 2. Table 3 EC50 values (blue dots, mean; blue lines, ± 1 SD of three assays) for the three active candidate compounds on one log-scaled nM axis. The solid vertical line marks 500 nM and the dashed line 1 μM. The orange markers show what the text says about each compound: below 500 nM for LabMol-149 and LabMol-152 (Abstract and final paragraph of the Discussion), and "submicromolar" for LabMol-151 (Discussion, third paragraph; that sentence also calls LabMol-149 submicromolar). LabMol-149's 3D7 interval lies entirely above 500 nM, and both LabMol-151 values lie above 1 μM. The W2 value for LabMol-149 sits at the threshold, and its ± 1 SD interval spans it.

Why it matters

The count of compounds meeting the potency criterion is the paper's headline experimental result. It appears in the Abstract ("two compounds ... (EC50 <500 nM)"), in the Author summary ("two new potential antimalarial candidates") and in the final paragraph of the Discussion, which says the two compounds "completely satisfy" the criteria for new antimalarial drugs. The Abstract also concludes that the approach found "two new families of potential next generation antimalarial agents, which are in compliance with the guidelines and criteria".

By the paper's own Table 3, one compound (LabMol-152, a quinazoline) meets the stated criterion. LabMol-149 fails it on the sensitive strain by a wide margin. Its 3D7 EC50 of 1450 nM is also above the 1 μM cut-off that the paper itself uses to label training compounds "active" (Methods, "Datasets": "1,058 active compounds with EC50 ≤ 1 μM"). On the resistant strain it is essentially at the threshold, at 509 nM. LabMol-149 remains an active and selective hit (SI above about 69 on 3D7 and 196 on W2), and nothing here questions the measurements themselves. But it is a low-micromolar hit, not a compound that satisfies the sub-500 nM criterion.

Readers who rely on the Abstract, for example to compile hit lists or to judge how well deep-learning virtual screening performs in antimalarial discovery, would count two qualifying compounds from two scaffold families where the data show one. The rest of the paper's conclusions, such as the QSAR models, the screen and LabMol-152's potency, do not depend on this. These are in vitro preclinical results, so there is no direct clinical consequence.

Explanations considered

Status

Suggested correction

  1. In the Abstract, change "two compounds, LabMol-149 and LabMol-152, showed potent antiplasmodial activity at low nanomolar concentrations (EC50 <500 nM)" so that it names LabMol-152 alone as meeting EC50 <500 nM. Describe LabMol-149 as active in the low-micromolar to high-nanomolar range (1.450 μM on 3D7, 0.509 μM on W2). Revise "two new families ... in compliance with the guidelines and criteria" and the Author summary's "two new potential antimalarial candidates" to match.
  2. In the final paragraph of the Discussion, restrict "completely satisfy the guidelines and criteria ... (EC50 <500nM)" to LabMol-152.
  3. In the third paragraph of the Discussion, replace "submicromolar" for LabMol-149 and LabMol-151, which are 1.450 / 0.509 μM and 1.911 / 1.616 μM, with "low micromolar" or the actual values.
  4. Minor: in the Results, "two compounds" is followed by three names; change it to "three compounds". Change LabMol-149's SIa from >71.4 to >69.0 (100/1.450), or state which EC50 value was used to compute it.